ChatWorks for Retail · Use case studies
The complete chat commerce channel, run by a few people, measured end to end.
White-label, AWS-hosted chat commerce for mid-to-large retailers in frontier and emerging markets. What shoppers see on WhatsApp, Telegram and web chat — and what a small operations team sees in the back office.
Chat, smartphones and mobile money are already the channel. Retail hasn't wired them together.
01 No chat commerce
A website with a cart, an app few download, a call centre. Shoppers with 'do you have this?' have nowhere to ask.
02 Human-centric chat
A WhatsApp number answered by a staff member from a handset — works for fifty conversations a day, collapses at five hundred, no record of orders or payments.
03 Support-only chat
A bot that answers opening hours and hands everything else to an agent — the channel shoppers open most is used for the least valuable task.
One channel, every step, from 'do you have it?' to 'did you like it?'
Shopper-facing, automated
- Product search and stock availability per branch
- Pricing, promotions and loyalty balance
- Cart management across sessions and channels
- Saved payment and delivery preferences
- Delivery tracking and delivery confirmation
- Returns, refunds and after-sales satisfaction
Back office, digitised
- Orders into a single queue tagged by channel and stage
- Payment processing (M-Pesa STK push, card, MoMo) with reconciliation
- Order-flow messaging at every stage change
- Dispatch handover (pick lists, rider assignment, proof of delivery)
- Stock signals back to branch and buying teams
- Disputes, refunds and loyalty adjustments in the same case
How it works
Shoppers talk to the retailer's own branded number and bot. ChatWorks handles the many-to-one dynamic: thousands of simultaneous inbound conversations, each carried through a defined stage machine. Operators see an inbox sorted by urgency, with a deal pane, timeline and internal notes per conversation, and step in only where a human adds value.
A chat commerce desk of six to twelve people running a channel that a call centre of a hundred could not.
A channel that measures itself.
Illustrative dashboard for a supermarket chain · not measured results
What is selling where at what margin
Who is buying and who isn't
Cohorts, loyalty, fleet accounts
The funnel and where it leaks
Abandoned carts by stage and reason, recoverable carts
How fast the back office moves
Medians and 90th percentiles by branch, shift, picker, rider
Service and returns
Complaints by type, refund rates, satisfaction per branch and zone
Channel and staff performance
Conversion by channel and campaign, share of orders with no human message, operator response/resolution
The retailers, people, conversations and figures in these case studies are illustrative composites created to show how a ChatWorks deployment would operate for each retail vertical. They do not describe any named company and are not measured results.
One AWS-native stack, white-labelled per retailer.
Channels
WhatsApp Business (Twilio messaging, template & session-window handling); Telegram Bot API webhook; embeddable web chat widget; one conversation record across all three.
Application & data
Next.js on AWS Amplify Hosting (SSR); DynamoDB for conversations, carts, orders, payments, catalogues, profiles; Cognito sign-in with MFA; S3 for images, receipts, booklist photos, proof of delivery; SQS/EventBridge for order events to dispatch, POS and ERP; SES for email.
Payments & integration
M-Pesa Daraja STK push with callback reconciliation; pawaPay for MoMo/Airtel Money; cards via Stripe; catalogue and stock via CSV upload or API sync from POS/ERP; optional escrow; CloudWatch monitoring, message error-rate alerts and a public status page.
Analytics
DynamoDB Streams → S3 data lake in the retailer's AWS account → Athena → QuickSight dashboards (funnel, demand, margin, operations timing, service and returns), tables shared with the retailer's BI team.
The operator console
Inbox sorted by urgency, filtered by channel and stage, with derived tags ('your turn', 'high value', 'repeat buyer', 'going cold'); deal pane with cart/job, timeline of every stage change and payment event, internal notes, quick-reply templates, WhatsApp session-window timer; stage transitions enforced by the platform (an order cannot be marked dispatched before it is paid).
What a deployment includes
- Tenant provisioning in the retailer's AWS account or ChatWorks-managed account per market
- Brand, numbers and bots registered in the retailer's name
- Catalogue, stock, branch, pricing and loyalty integration
- Payment provider onboarding and reconciliation reports
- Retail intelligence dashboards
- Conversation design for the retailer's vertical in English, Swahili and French
- Operator training, playbooks and a go-live period with ChatWorks staff on the desk
Talk to us about your channel.
View demo
Digital clickthrough interactive demo.
If your retail business uses chat today with a person on the other end, or not at all, we can show a working demo of your catalogue on WhatsApp, Telegram and web chat within a week.